amie.mining

Class AMIE

java.lang.Object

amie.mining.AMIE

public class AMIE
extends java.lang.Object

Main class that implements the AMIE algorithm for rule mining on ontologies. The ontology must be provided as a list of TSV files where each line has the format SUBJECT<TAB>RELATION<TAB>OBJECT.

Author:

lgalarra

Constructor Summary

Constructors

Constructor and Description
AMIE(MiningAssistant assistant, int minInitialSupport, double threshold, Metric metric, int nThreads)

Method Summary

Modifier and Type	Method and Description
MiningAssistant	getAssistant()
static AMIE	getInstance(java.lang.String[] args) Gets an instance of AMIE configured according to the command line arguments.
static AMIE	getLossyInstance(KB db, double minPCAConfidence, int minSupport) Returns an instance of AMIE that enables the lossy optimizations, i.e., optimizations that optimize for runtime but that could in principle omit some rules that should be mined.
static AMIE	getLossyVanillaSettingInstance(KB db, double minPCAConfidence, int startSupport) Returns an (vanilla setting) instance of AMIE that enables the lossy optimizations, i.e., optimizations that optimize for runtime but that could in principle omit some rules that should be mined.
Metric	getPruningMetric()
java.util.Collection<javatools.datatypes.ByteString>	getSeeds()
static AMIE	getVanillaSettingInstance(KB db) Returns an instance of AMIE that mines rules on the given KB using the vanilla setting of head coverage 1% and no confidence threshold.
static AMIE	getVanillaSettingInstance(KB db, double minPCAConfidence) Returns an instance of AMIE that mines rules on the given KB using the vanilla setting of head coverage 1% and a given PCA confidence threshold
boolean	isRealTime()
boolean	isVerbose()
static void	main(java.lang.String[] args) AMIE's main program
java.util.List<Rule>	mine() The key method which returns a set of rules mined from the KB based on the AMIE object's configuration.
void	setCheckParentsOfDegree2(boolean booleanVal)
void	setPruningMetric(Metric pruningMetric)
void	setRealTime(boolean realTime)
void	setSeeds(java.util.Collection<javatools.datatypes.ByteString> seeds)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

AMIE

public AMIE(MiningAssistant assistant,
            int minInitialSupport,
            double threshold,
            Metric metric,
            int nThreads)

Parameters:

assistant - An object that implements the logic of the mining operators.

minInitialSupport - If head coverage is defined as pruning metric, it is the minimum size for a relation to be considered in the mining.

threshold - The minimum support threshold: it can be either a head coverage ratio threshold or an absolute number.

metric - Head coverage or support.

Method Detail

getAssistant

public MiningAssistant getAssistant()

isVerbose

public boolean isVerbose()

isRealTime

public boolean isRealTime()

setRealTime

public void setRealTime(boolean realTime)

getSeeds

public java.util.Collection<javatools.datatypes.ByteString> getSeeds()

setSeeds

public void setSeeds(java.util.Collection<javatools.datatypes.ByteString> seeds)

getPruningMetric

public Metric getPruningMetric()

setPruningMetric

public void setPruningMetric(Metric pruningMetric)

mine

public java.util.List<Rule> mine()
                          throws java.lang.Exception

The key method which returns a set of rules mined from the KB based on the AMIE object's configuration.

Returns:

Throws:

java.lang.Exception

setCheckParentsOfDegree2

public void setCheckParentsOfDegree2(boolean booleanVal)

getVanillaSettingInstance

public static AMIE getVanillaSettingInstance(KB db)

Returns an instance of AMIE that mines rules on the given KB using the vanilla setting of head coverage 1% and no confidence threshold.

Parameters:

db -

Returns:

getVanillaSettingInstance

public static AMIE getVanillaSettingInstance(KB db,
                                             double minPCAConfidence)

Returns an instance of AMIE that mines rules on the given KB using the vanilla setting of head coverage 1% and a given PCA confidence threshold

Parameters:

db -

Returns:

getLossyVanillaSettingInstance

public static AMIE getLossyVanillaSettingInstance(KB db,
                                                  double minPCAConfidence,
                                                  int startSupport)

Returns an (vanilla setting) instance of AMIE that enables the lossy optimizations, i.e., optimizations that optimize for runtime but that could in principle omit some rules that should be mined.

Parameters:

db -

minPCAConfidence -

startSupport -

Returns:

getLossyInstance

public static AMIE getLossyInstance(KB db,
                                    double minPCAConfidence,
                                    int minSupport)

Returns an instance of AMIE that enables the lossy optimizations, i.e., optimizations that optimize for runtime but that could in principle omit some rules that should be mined.

Parameters:

db -

minPCAConfidence -

minSupport -

Returns:

getInstance

public static AMIE getInstance(java.lang.String[] args)
                        throws java.io.IOException,
                               java.lang.InstantiationException,
                               java.lang.IllegalAccessException,
                               java.lang.IllegalArgumentException,
                               java.lang.reflect.InvocationTargetException

Gets an instance of AMIE configured according to the command line arguments.

Parameters:

args -

Returns:

Throws:

java.io.IOException

java.lang.reflect.InvocationTargetException

java.lang.IllegalArgumentException

java.lang.IllegalAccessException

java.lang.InstantiationException

main

public static void main(java.lang.String[] args)
                 throws java.lang.Exception

AMIE's main program

Parameters:

args -

Throws:

java.lang.Exception